The recent India–AI Impact Summit 2026 demonstrated a defining global inflection point — the transition from dialogue to demonstrable impact when it comes to artificial intelligence (AI). Anchored in the principles of People, Planet, and Progress, it envisioned a future where AI advances humanity, fosters inclusive growth, and safeguards our shared planet.
With just a few minutes of recorded audio, attackers can create convincing voice replicas capable of deceiving even close acquaintances.
AI has reshaped the digital landscape, offering powerful tools for innovation but also new avenues for threats. According to ISACA’s 2026 Tech Trends and Priorities Pulse Poll, only 13 percent of global respondents indicated that their organization is very prepared to manage the risks associated with generative AI solutions this year, with 30 percent saying they are not very or not at all prepared.
One emerging threat is AI voice cloning, a technique that uses machine learning (ML) to replicate a person’s voice with alarming precision. While this AI / ML technology has legitimate applications in many industries, it is also weaponized by threat actors in increasingly sophisticated social engineering attacks.
AI voice cloning involves training a model on audio samples of a person’s speech to generate synthetic voice outputs that mimic their tone, cadence, and inflection. With just a few minutes of recorded audio, attackers can create convincing voice replicas capable of deceiving even close acquaintances.
The scam starts with a 5-10 second voice sample. The scammers capture it from sources such as:
The scammers create a sense of panic with an extremely emotional scene. It could be a case of a road accident and then seeking for immediate funds transfer for hospitalization or it could be a case of kidnapping and then seeking for ransom or any so-called emergency situation. They also often create background sounds, such as someone crying for help, a fake sniffle, sounds that convey urgency, or an instance of a kidnapper shouting instructions. They add the sounds so well such that it overrides logical reasoning.
They call using spoofed phone numbers so that their identity is hidden. These numbers often show up as spoofed caller IDs, international calls or even local mobile numbers. They play the cloned voice at the right emotional peak (a child crying, the parent begging, the emotional and urgent – “please, help me”). The logical reasoning takes a back seat while the emotional drama sets in.
The scammers demand instant digital payment in the form of Quick Response (QR) code payment, Unified Payment Interface (UPI) payment or even via wallet payment.
After payment is made, they cut the call and disappear. They are nowhere to be found.
By the time the victim realises it as a cybercrime, the money is lost and laundered across multiple accounts.
Over the years, AI-voice clone scams have increased multifold. AI-generated voices that impersonate people have increased in quality far beyond what even many experts expected would be the case just a few years ago. People tend to forget that voice is a personally identifiable information (PII) that needs to be protected.
Research from McAfee found that just three to four seconds of audio can create a voice clone with an 85% accuracy match. Twenty percent of people who were surveyed in India confirmed that they experienced an AI-voice clone scam and 27% responded that it happened to someone they knew. These statistics were reported to be the highest in India as compared to other countries, while US stood at the second highest and UK, the third highest.
The report highlights an interesting statistic on the situations to which Indians would respond to the voicemail or voice note from a family member or friend. Sixty-nine percent responded to car issues or accident, 70% of them responded to a voice call creating a situation of a theft victim, 65% responded to lost wallet and phone scenarios and 62% responded to a case in which someone claimed to be on a vacation abroad and needed help. Again, these numbers were reportedly the highest for India when compared to other countries. Further, 38% of Indians could not tell the difference between a voicemail from a loved one and one that is AI-generated.
As Fortune reported in December 2025, voice cloning has crossed the “indistinguishable threshold” — meaning human listeners can no longer reliably distinguish cloned voices from authentic ones. A few seconds of audio now suffice to generate a convincing clone – complete with natural intonation, rhythm, emphasis, emotion, pauses and breathing noise.
If you are not vigilant, 2026 will be the year you get fooled by a deepfake! However, there are some steps that people can take to protect themselves, including:
Guest author Chetan Anand is the Associate Vice President of Information Security and CISO at Profinch Solutions and ISACA Global Mentor and Emerging Trends Working Group Member. Any opinions expressed in this article are strictly those of the author.
The Tech Panda looks at Indian Fintech expanding to Sri Lankan Bank while a Swiss…
By enabling secure, consent-based financial data sharing, the Account Aggregator framework is laying the groundwork…
There’s a quiet crisis in one of medicine’s most exciting fields. Cell therapy – the…
Lenovo and NVIDIA are pushing AI into its next phase, scaling real-time, production-ready systems that…
AI is changing how work gets done. It’s also changing how people feel about their…
The Tech Panda takes a look at recent funding events in the tech ecosystem, seeking…